Web Mapping & Visualisation

The Web’s Architecture and Economy

Matthew Howard, Elisabetta Pietrostefani & Daniel Arribas Bel

Today

  • A (brief an opinionated) history of the Web
  • The server/client model
  • The modern web mapping eco-system
  • API overview

Pre 1970s

The seeds:

  • US (e.g. Licklider’s “Galactic Network”)

  • Mostly military contracts (e.g. D/ARPA –> ARPANET) and “research’y”

  • Develop protocols for machine communication

1970s - Birth of the internet

centered image

1980s

  • Growth of the “web”

  • From experimental validation to scaled up insfrastructure

  • Free software (e.g. “Free as in Freedom”)

1990s

  • Civilian and commercial growth

  • Web 1.0

  • Open Source software (e.g. “The cathedral and the Bazaar”)

2000s

  • Web 2.0

  • Mobile

  • Web mapping takes off (hello Google Maps!)

2010s

  • Consolidation of ‘GAFA’ –> concentration

  • IoT

  • Death of the desktop?

2020s

  • Web3

  • Government regulation and legislation

  • AI

Ideas to retain

  • The Web is technology to build decentralised systems

  • Economics (for the most part) have turned it into a concentrated economy

  • Computing today is physically distributed but socio-economically concentrated

The server/client model

centered image

Benefits

  • Interoperability of disparate platforms

  • Optimise on hard/software for each task (“distribute”)

  • Separate data collection (e.g. sensor), storage (e.g. data centre), intensive computing (e.g. compute cluster), interaction (e.g. mobile)

Disadvantages

  • Requires (cheap & ubiquitous) connectivity

  • More complex than an isolated approach (e.g. desktop)

  • Harder to “keep afloat”

Building blocks of a web map

centered image

The current web mapping landscape

Structure

  • Software: a lot of open-source projects

  • Platforms: a concentrated few (web infrastructure is hard and expensive!)

  • Business model: software as a service

The trade off…

centered image

This course: Not focused on engineering the backend/infastructure, prioritises frontend design/development

Data challenges

  • Bias (Who is/isn’t represented?)

  • Licensing (Who controls the data, what can you do with it?)

  • Access (How technically complex is to use?)

APIs

What do APIs actually do?

  • Application Programming Interfaces (“APIs”)
  • Instead of having to download a data set, APIs allow you to request (parts of) a database directly from a remote server to a local machine.

When you work with web APIs, two different computers - a client and server - will interact with each other to request and provide data, respectively.

centered image

Web API structure

  • Often require Authentication using tokens (potentially linked to billing)

  • Adhere to a particular style known as Representation State Transfer or REST (in most cases)

  • RESTful APIs are convenient because we use them to query database using URLs over HTTP

Accessing APIs

Plug-n-play packages. Many common APIs are available through user-written Py libraries.

Writing our own API request. If no wrapper function is available, we have to write our own API request and format the response ourselves.

Week 3 Lab Content

Main Goal: Testing out example APIs and access methods:

  • ‘Plug-n-Play’ API packages with the Census package.

  • Writing our own requests looking at bike sharing in London.

  • Spatial data APIs (OSM data using the geocoder from geopy)

Creative Commons License
Web Mapping and Visualisation by Elisabetta Pietrostefani is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.